We will conduct a series of statistical analyses on four large data sets collected by the Principal Investigator and his associates as part of previously funded NIMH projects (N = 7,218 who completed questionnaire; 3,760 who were diagnostically interviewed). Two data sets are based on people over the age of 50; one data set is based on older adolescents (14- 18 years of age). The data incorporate measures for (a) psychopathology, including all past and current episodes of mental disorder; (b) existence of past and present physical illnesses, symptoms, and complaints; (c) mental and other health service utilization; (d) social status attributes; (e) family history of psychopathology; (f) cognitions; (g) coping skills; (h) aspects of personality functioning; (i) interpersonal interaction; (j) social support; (k) stress (macro and micro); (l) independent living skills; and (m) aspects of neuropsychological and psychophysiological functioning. All data have been entered, edited, and are accessible to use in further analyses. Important questions about depression can be addressed and answered at relatively low cost. Issues we will address in this proposal include the following: 1. The impact of psychosocial risk factors, singly and in combination, on the occurrence of depression. 2. The ability of psychosocial risk factors to predict depression as a function of age at onset, gender, and number of previous episodes of depression; the risk factors for clinical and non-clinical depression. 3. The development and evaluation of theoretical models for the occurrence and maintenance of depression. Using results from our primary analyses and capitalizing on the two-panel nature of the data sets, we will develop and test specific theoretical models using Structural Equation Modeling (SEM). 4. The nature of relations between psychosocial variables and depression. to what extent are they state-specific? Do the temporal relations with depression vary as a function of gender or age? 5. The operating characteristics of the Center for Epidemiologic Studies Depression Scale (CES-D) as a screener for depression when used with older adults. These analyses will be conducted in collaboration with a group of consultants each of whom is at the cutting edge of research in an area of depression directly relevant to the proposal. All of our consultants have data sets, which will allow for cross-validation of findings.